Qwen2_CN_NSFW_GGUF needs about 5.7 GB of memory at Q4_K_M (a 4.7 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 2 quants, from 4.7 GB (Q4_K_M) to 15.2 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.9 GB.
Qwen2_CN_NSFW_GGUF – 7.62B multimodal model fine‑tuned for Chinese content
Qwen2_CN_NSFW_GGUF is a 7.62B parameter vision‑language model derived from Qwen2.5‑VL‑7B‑Instruct. It supports image‑to‑text tasks and can process Chinese text. The model is released under the apache‑2.0 license and is available in two GGUF quantizations.
Good at
- Image‑to‑text generation
- Chinese language understanding
- Vision‑language reasoning
Limits
- No vision support beyond image‑to‑text
- Requires 8k context for memory estimates
- No explicit vision‑language benchmarks provided
Best for
- Chatbots needing Chinese image understanding
- Document analysis with Chinese text
- Multimodal content moderation
Who it’s for. Developers building Chinese multimodal applications who need a lightweight, open‑source model. The model is suitable for local deployment with llama.cpp or similar runtimes.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 4.7 GB | 5.7 GB | Fits | Get file → | |
| F16 | 15.2 GB | 16.2 GB | Too big | Get file → |
For 8 GB VRAM, use the Q4_K_M quantization (file size 4.68 GB, estimated memory 5.7 GB at 8k context). For higher precision and more VRAM, use the F16 quantization (file size 15.24 GB, estimated memory 16.2 GB at 8k context).
What GPU do you need for Qwen2_CN_NSFW_GGUF?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: the best quant that fits is Q4_K_M (4.7 GB file, about 5.7 GB in use).
- 8 GB VRAM: the best quant that fits is Q4_K_M (4.7 GB file, about 5.7 GB in use).
- 12 GB VRAM: the best quant that fits is Q4_K_M (4.7 GB file, about 5.7 GB in use).
- 16 GB VRAM: the best quant that fits is Q4_K_M (4.7 GB file, about 5.7 GB in use).
- 24 GB VRAM: the best quant that fits is F16 (15.2 GB file, about 16.2 GB in use).
Which Qwen2_CN_NSFW_GGUF quant should you choose?
Q4_K_M is the usual starting point for Qwen2_CN_NSFW_GGUF: it balances file size and quality well. Here it is a 4.7 GB file.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Qwen2_CN_NSFW_GGUF
How much VRAM does Qwen2_CN_NSFW_GGUF need?
About 5.7 GB at Q4_K_M with an 8K context window. That is the 4.7 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Qwen2_CN_NSFW_GGUF on an 8 GB GPU?
Yes, with Q4_K_M: the 4.7 GB file needs about 5.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Qwen2_CN_NSFW_GGUF GGUF?
It depends on the quant: Q4_K_M is 4.7 GB, F16 is 15.2 GB.
What context length does Qwen2_CN_NSFW_GGUF support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.5 GB at 8K and 1.9 GB at 32K.
Where can I download Qwen2_CN_NSFW_GGUF GGUF?
The files are on Hugging Face in the WSDW/Qwen2_CN_NSFW_GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Qwen2_CN_NSFW_GGUF use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Qwen2_CN_NSFW_GGUF on 8 GB VRAM?
Yes, the Q4_K_M quantization requires approximately 5.7 GB of VRAM at 8k context, which fits within an 8 GB budget.
Does the model support English text?
The parent model supports English, but the fine‑tuned version focuses on Chinese content; English support is not guaranteed.
Is the model suitable for vision‑language tasks beyond image‑to‑text?
The model is primarily designed for image‑to‑text generation and does not include advanced vision‑language reasoning beyond that scope.